Report by NewsroomPlus.com Contributed by Ben K Jarman, J.P. Morgan Australia Limited
It has been said that economists make weather forecasters look good.
Combining the two therefore seems especially dangerous, but it is nevertheless worth mentioning that the experts are flagging the likelihood of a strong El Nino event this year. In New Zealand, El Ninos have played out as negative supply shocks to the primary sector, with floods in normally dry areas, and drought in normally fertile areas.
As an agricultural commodity exporter, El Nino depresses real primary sector output, real exports, and real GDP growth in New Zealand, but typically raises export prices, and has mixed effects on inflation. In general the implications for the currency are therefore not obvious, but at this juncture we see the RBNZ being most sensitive to incremental growth weakness, and the mix of outcomes as worsening an already deteriorating external balance. Both should make an El Nino event negative for NZD, consistent with our forecast bias (NZD/USD to 0.58 over the next 12 months).
The 2015 El Nino
New Zealand’s weather authority, MetService, noted this month that the 2015 El Nino event “has reached maturity in the tropical Pacific Ocean. Both the ocean and atmosphere have surpassed strong thresholds.” Sea surface temperatures are a couple of degrees warmer than normal, and the Southern Oscillation Index (SOI, see Figure 1) is deep into negative (i.e. El Nino) territory, which hasn’t happened to that degree since the severe event of 1997-98. Rainfall in New Zealand is only slightly below average at this stage, but the El Nino phenomenon is expected to strengthen into year-end, with the potential to drive a much drier than normal summer and early autumn, which depresses winter crop yields and pastures for livestock/dairy into mid next year.
Below we have used the meteorological bureau’s classification of El Nino events over the last 20 years (1994, 1997-98, 2002-03, 2006, 2009-10) to examine how the phenomenon typically affects macro variables and NZD/USD. At first glance there does not appear to be any obvious pattern in terms of how El Nino affects the currency (Figure 2).
But it is important to first control for other dynamics playing out in the economy to find the marginal influence of this supply shock. We model a vector of macro variables as a system, adding in a dummy variable to represent the effect of El Nino periods (more details at the bottom). The results discussed below therefore represent the ‘average’ El Nino effect after controling for other influences.
Comparing our own results with other studies, the most convincing finding is the effect of El Nino on primary sector GDP, which, after incorporating that sector’s share of total output, reduces real GDP growth by a total 0.5%-pts over 4 quarters (Figure 3). This is in the ballpark of the 0.3%-pt drag estimate made by the IMF (see here) and the 0.3-0.6%-pt drag the RBNZ pegged on the 2013 drought (see here). The RBNZ’s work is in particular worth mentioning as, unlike our approach which treats all El Ninos as equal, their work calibrates the extent of dry weather conditions to the geography and the time of year when it will do most damage. An El Nino that creates dry weather through March, as in the 2013 drought, would therefore push estimates of the growth drag to the upper end of the above range.
What muddies the waters somewhat with El Nino is the fact that the negative supply shock to the primary sector typically pushes up export prices, on our estimates by 4-5% on average, creating some complications for inflation and monetary settings. Certainly the 2013 drought (not technically an El Nino episode) prompted a steep rise in dairy prices and made the RBNZ turn hawkish. However, our sense is that with dairy prices having fallen sharply over the past 18 months (even after the bounce of the last few auctions), and with the RBNZ having just delivered deep cuts to their GDP forecasts, the central bank will be more sensitive to the real GDP growth hit than to an export price rally off the lows.
Another of our findings is that El Nino events are on average quite inflationary (+0.9%-pts). The RBNZ study mentioned above finds mixed effects (e.g. food prices up, non-tradables inflation down), while the IMF finds New Zealand to be an outlier among Southern Hemisphere economies, with El Nino being deflationary. The IMF concede, however, that this finding may be picking up the broader trend toward lower inflation through the sample, as inflation targeting took hold. We do not view our finding of inflationary El Ninos as having great significance for monetary policy, given the persistent undershooting of the inflation target in recent years. The fact that the inflation effect is measured to be positive, and similarly that NZD has historically risen more often than fallen during El Nino episodes, likely reflects the fact that the output gap happened to be positive on average when these events started. At the moment, the output gap is slightly negative, and GDP growth is moderating, which tilts the balance of risks to medium term inflation from an El Nino shock to the downside (Figure 4).
Risks still biased to weaker news on the external sector
We have been bearish NZD on the basis of the dairy sector’s weakness, the RBNZ’s change in stance, and external vulnerabilities. Tempering that to an extent has been our constructive view on the New Zealand household sector’s ability to withstand the income shock from the agriculture sector.
The information to date on consumer spending and the housing market still suggests resilience. However, as the size of the primary sector shock gets bigger, it represents a bigger hit to headline GDP, in a world where the central bank is proving very sensitive to downside risks. As RBNZ Governor Wheeler made clear in last week’s MPS press conference, El Nino is on his radar too – Wheeler mentioned El Nino, alongside a “dramatic” China slowdown, as the most likely catalysts of recession in New Zealand that would bring “substantial” further rate cuts. Even if things don’t get that serious, the combination of deteriorating primary sector exports and stable domestic demand growth under an El Nino event would push the current account deficit wider, and increase the call on foreign savings at a moment when markets are losing patience with current account deficit nations.
Model details: We estimate a VAR system, using quarterly data with 4 lags on; global GDP growth; the NZ output gap (2 qtr lag); NZ export price growth (USD terms); the quarterly change in NZD/USD; 1 quarter change in the RBNZ OCR; primary sector GDP growth; and quarterly inflation. We then add an exogenous dummy variable set to 1 for quarters in which El Nino was active. Alternative specifications involved including the El Nino dummy as an endogenous variable, placed first in the Cholesky ordering, with other variables as above. This did not change the estimates materially.
Source provided to firstname.lastname@example.org: markets.jpmorgan.com/research